1: % LaTeX template file for
2: % Publications of the Astronomical Society of Australia.
3: % Version 2.6 - 5 April 2001
4: %
5: % The most current version of this file can be found at:
6: % http://www.atnf.csiro.au/pasa/template
7: %
8: % Other information on Electronic PASA can be found at:
9: % http://www.atnf.csiro.au/pasa/
10: %
11: % Instructions for submitting to Electronic PASA can be found at:
12: % http://www.atnf.csiro.au/pasa/authors.html
13: %
14: % General instruction for PASA authors can be found at:
15: % http://www.atnf.csiro.au/asa_www/instructions.html
16: %
17: % Do not change any arguments to LaTeX commands unless the argument is
18: % written in ALL CAPITALS as a place-holder, or instructed otherwise by
19: % preceding comments.
20: %
21: % Style files psfig, epsf & html are acceptable. Do not use any other
22: % style files.
23: % Point size of text may be changed if desired.
24: %
25: % \label and \ref cross-referencing is encouraged for referencing
26: % figures and sections - these will appear as hyperlinks in the HTML version.
27: %
28: %\documentstyle[12pt,psfig]{article}
29: \documentstyle[gcdv,psfig]{article}
30: %
31: % Baselineskip may be altered if desired.
32: %
33: \baselineskip=2em
34: %
35: % A few definitions. Do not change the reference command.
36: %
37: \def\reference{\parskip 0pt\par\noindent\hangindent 0.5 truecm}
38: \def\kms{km ${\rm s}^{-1}$}
39: %
40: % Text locations - these may be altered slightly if desired.
41: %
42: %\textwidth=15.5cm
43: %\textheight=22.6 cm
44: %\topmargin=-1.0 cm
45: %\oddsidemargin=0.5cm
46: %\evensidemargin=0.5cm
47:
48: \textwidth=17cm
49: \textheight=24.5 cm
50: \topmargin=-1.0 cm
51: \oddsidemargin=-0.5cm
52: \evensidemargin=-0.5cm
53:
54: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
55: % OWN MACROS
56: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
57:
58: \def\lesssim{{_ <\atop{^\sim}}}
59: \def\grtsim{{_ >\atop{^\sim}}}
60: \def\ap3m{AP$^3$M}
61: \def\LCDM{$\Lambda$CDM}
62: \def\LWDM{$\Lambda$WDM}
63: \def\hkpc{$h^{-1}{\ }{\rm kpc}$}
64: \def\hMpc{$h^{-1}{\ }{\rm Mpc}$}
65: \def\hMsun{$h^{-1}{\ }{\rm M_{\odot}}$}
66: \def\kms{${\rm{\ }km{\ }s^{-1}}$}
67: \def\nbody{$N$-body}
68: \def\c15{$c_{\rm 1/5}$}
69: \def\llink{$b_{\rm link}$}
70: \def\rsmooth{$r_{\rm smooth}$}
71: \def\rmhf{$r_{\tt MHF }$}
72: \def\Rvir{$R_{\rm vir}$}
73: \def\Mvir{$M_{\rm vir}$}
74: \def\rvir{$r_{\rm vir}$}
75: \def\vvir{$v_{\rm vir}$}
76: \def\rsat{$r_{\rm sat}$}
77: \def\Dsat{$D_{\rm sat}$}
78: \def\msat{$m_{\rm sat}$}
79: \def\vsat{$v_{\rm sat}$}
80: \def\rtrunc{$r_{\rm trunc}$}
81: \def\rtidal{$r_{\rm tidal}$}
82: \def\rmax{$r_{\rm max}$}
83: \def\zform{$z_{\rm form}$}
84: \def\vecx{\vec{x}}
85: \def\vecp{\vec{p}}
86: \def\vecr{\vec{r}}
87: \def\vecF{\vec{F}}
88: \def\vecg{\vec{g}}
89: \def\ddx{\ddot{\vec{x}}}
90: \def\ddr{\ddot{{r}}}
91: \def\dx{\dot{\vec{x}}}
92: \def\vecnabla{\vec{\nabla}}
93: \def\dvecx{\dot{\vec{x}}}
94: \def\ddvecx{\ddot{\vec{x}}}
95: \def\ddvecr{\ddot{\vec{r}}}
96: \def\REF{\textbf{REFS}}
97: \newcommand{\Table}[1]{Table~\ref{#1}}
98: \newcommand{\Sec}[1]{Section~\ref{#1}}
99: \newcommand{\Eq}[1]{Eq.~(\ref{#1})}
100: \newcommand{\Fig}[1]{Figure~\ref{#1}}
101: \newcommand{\mlapm}{\texttt{MLAPM}}
102: \newcommand{\mhf}{\texttt{MHF}}
103: \newcommand{\mht}{\texttt{MHT}}
104: \newcommand{\GKGI}{\textsf{GKG}\textrm{I}}
105: \newcommand{\GKGII}{\textsf{GKG}\textrm{II}}
106: \newcommand{\GKGIII}{\textsf{GKG}\textrm{III}}
107:
108: \def\ea{et~al.~} % \ea = et al.
109: \def\u{\"u} % \u = "u
110: \def\o{\"o} % \o = "o
111: \def\a{\"a} % \a = "a
112:
113: \def\lesssim{\mathrel{\hbox{\rlap{\hbox{\lower4pt\hbox{$\sim$}}}\hbox{$<$}}}}
114: \def\gtrsim{\mathrel{\hbox{\rlap{\hbox{\lower4pt\hbox{$\sim$}}}\hbox{$>$}}}}
115:
116:
117: \newcommand{\AN}[3] {\mbox{AN~\textbf{#1},~#2~(#3)}}
118: \newcommand{\AAA}[3] {\mbox{A\&A~\textbf{#1},~#2~(#3)}}
119: \newcommand{\AAR}[3] {\mbox{A\&A~Rev.~\textbf{#1},~#2~(#3)}}
120: \newcommand{\AAS}[3] {\mbox{A\&A~Suppl.~\textbf{#1},~#2~(#3)}}
121: \newcommand{\ApJ}[3] {\mbox{ApJ~\textbf{#1},~#2~(#3)}}
122: \newcommand{\ApJS}[3] {\mbox{ApJ~Suppl.~\textbf{#1},~#2~(#3)}}
123: \newcommand{\ApJL}[3] {\mbox{ApJ~Lett.~\textbf{#1},~#2~(#3)}}
124: \newcommand{\ARAA}[3] {\mbox{Ann.~Rev.~A~\&~A~\textbf{#1},~#2~(#3)}}
125: \newcommand{\AJ}[3] {\mbox{Astron.~J.~\textbf{#1},~#2~(#3)}}
126: \newcommand{\MNRAS}[3] {\mbox{MNRAS~\textbf{#1},~#2~(#3)}}
127: \newcommand{\Nature}[3] {\mbox{Nature~\textbf{#1},~#2~(#3)}}
128: \newcommand{\NewA}[3] {\mbox{NewA~\textbf{#1},~#2~(#3)}}
129: \newcommand{\Science}[3]{\mbox{Science~\textbf{#1},~#2~(#3)}}
130: \newcommand{\PhRevL}[3] {\mbox{Phys.~Rev.~Lett.~\textbf{#1},~#2~(#3)}}
131: \newcommand{\PhRevD}[3] {\mbox{Phys.~Rev.~\textbf{D#1},~#2~(#3)}}
132: \newcommand{\astroph}[1]{\mbox{\texttt{astro-ph/#1}}}
133:
134:
135: %
136: % Start of document
137: %
138: \begin{document}
139:
140: %%%%%%%%%%% Emulate PASA style %%%%%%%%%%%%%%%%%5
141: \small
142: \shorttitle{Interactions of Satellite Galaxies}
143: \shortauthor{A. Knebe~\ea}
144: %%%%%%%%%%% End Emulate PASA style %%%%%%%%%%%%%%%%%5
145:
146: %
147: % Title
148: % Capitalise the title normally - do not use ALL CAPS.
149: %
150: \title
151: %%%%%%%%%%% Emulate PASA style %%%%%%%%%%%%%%%%%5
152: {\large \bf
153: %%%%%%%%%%% Emulate PASA style %%%%%%%%%%%%%%%%%5
154: %{
155: Interactions of Satellite Galaxies in Cosmological Dark Matter Halos
156: }
157: %
158:
159: % Authors
160: % Here comes the author(s) of the paper. Please add the appropriate author
161: % names for your paper and indicate within the $^...$ the number(s)
162: % which corresponds to the institute(s) of each author. In this example
163: % the second author has two institutional affiliations.
164: % Add or remove authors as required.
165: % **** IMPORTANT: Leave the closing curly bracket line as is. ******
166:
167: %%%%%%%%%%% Emulate PASA style %%%%%%%%%%%%%%%%%5
168: \author{\small
169: %%%%%%%%%%% Emulate PASA style %%%%%%%%%%%%%%%%%5
170: %\author{
171: Alexander Knebe $^{1}$, Stuart P.D. Gill, Brad K. Gibson
172: } % IMPORTANT: leave this curly bracket as the first character of this line.
173:
174: % Date - leave this blank.
175: \date{}
176: %%%%%%%%%%% Emulate PASA style %%%%%%%%%%%%%%%%%
177: \twocolumn[
178: %%%%%%%%%%% Emulate PASA style %%%%%%%%%%%%%%%%%5
179: \maketitle
180: \vspace{-20pt}
181: \small
182: % Institutions
183: % Here fill in your institute name(s) and address(es)
184: % The number in $^...$ indicates the author number. For example
185: {\center
186: Centre for Astrophysics \& Supercomputing,
187: Swinburne University, Mail \#31, P.O. Box 218, Hawthorn, VIC 3122,
188: Australia\\
189: $^1$aknebe@astro.swin.edu.au\\[3mm]
190: }
191:
192: % Abstract
193: % Simply place your abstract between the \begin{abstract} and
194: % \end{abstract} commands.
195: %
196: %\begin{abstract}
197: %%%%%%%%%%% Emulate PASA style %%%%%%%%%%%%%%%%%
198: \begin{center}
199: {\bfseries Abstract}
200: \end{center}
201: \begin{quotation}
202: \begin{small}
203: \vspace{-5pt}
204: %%%%%%%%%%% Emulate PASA style %%%%%%%%%%%%%%%%%
205: % Place the abstract here.
206: We present a statistical analysis of the interactions between
207: satellite galaxies in cosmological dark matter halos taken from fully
208: self-consistent high-resolution simulations of galaxy clusters. We
209: show that the number distribution of satellite encounters has a tail
210: that extends to as many as 3--4 encounters per orbit. On average 30\%
211: of the substructure population had at least one encounter (per orbit)
212: with another satellite galaxy. However, this result depends on the age
213: of the dark matter host halo with a clear trend for more interactions
214: in younger systems. We also report a correlation between the number
215: of encounters and the distance of the satellites to the centre of the
216: cluster: satellite galaxies closer to the centre experience more
217: interactions. However, this can be simply explained by the radial
218: distribution of the substructure population and merely reflects the
219: fact that the density of satellites is higher in those regions.
220:
221: In order to find substructure galaxies we applied (and present) a new
222: technique based upon the \nbody\ code \mlapm. This new halo finder
223: \mhf\ (\mlapm's-halo-finder) acts with exactly the same accuracy as
224: the \nbody\ code itself and is therefore free of any bias and spurious
225: mismatch between simulation data and halo finding precision related to
226: numerical effects.
227:
228: %%%%%%%%%%% Emulate PASA style %%%%%%%%%%%%%%%%%
229: %\end{abstract}
230: %%%%%%%%%%% End Emulate PASA style %%%%%%%%%%%%%%%%%
231: {\bf Keywords: methods: $n$-body simulations -- galaxies: clusters --
232: galaxies: kinematics and dynamics -- cosmology: dark matter
233: }
234: %%%%%%%%%%% Emulate PASA style %%%%%%%%%%%%%%%%%
235: \end{small}
236: \end{quotation}
237: ]
238: %%%%%%%%%%% End Emulate PASA style %%%%%%%%%%%%%%%%%
239:
240: % Place keywords here. Please write all keywords in lower case. PASA uses the
241: % standard list of subject
242: % headings adopted by The Astrophysical Journal and available from URL:
243: % http://www.journals.uchicago.edu/ApJ/keywords_text.html
244:
245: % A formatting command to add space between the author list and the body
246: % of the paper when printed. This spacing may be changed as desired.
247: \bigskip
248:
249:
250: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
251: \section{Introduction}
252: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
253: \subsubsection*{Observations}
254: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
255: There are several hints indicating that satellite galaxies orbiting
256: within our own Milky Way are interacting with each other. Zhao
257: (1998), for instance, proposed a scenario where the Sagittarius Dwarf
258: galaxy had an encounter with the Magellanic Cloud system some 2--3
259: Gyrs ago, something that has also been speculated and noted by
260: Ibata~\& Lewis (1998). Moreover, the two Magellanic Clouds themselves
261: are another example of an interacting pair of substructure
262: galaxies. It has also been noted by Moore~\ea (1996) that ``galaxy
263: harrasment'' in cosmological simulations of galaxy cluster evolution
264: will lead to a morphology change of satellite galaxies.
265:
266: However, the literature to date lacks a statistical analysis of
267: interacting satellite galaxies orbiting within the potential of a
268: common dark matter host halo. How frequent are satellite-satellite
269: encounters and where in the galaxy cluster do they happen?
270: Furthermore, observations of the Local Group Dwarfs indicate a clear
271: correlation between star formation activity and the distance of the
272: respective Dwarf to the centre of the Milky Way (van den Bergh 1994)
273: with satellites farther away showing stronger activity. Can this be
274: ascribed to satellite-satellite interactions? The aim of this study is
275: to quantify such interactions in galaxy clusters derived from fully
276: self-consistent cosmological \nbody\ simulations within the framework
277: of the currently accepted Cold Dark Matter (CDM) structure formation
278: scenario.
279:
280: \subsubsection*{Is Cold Dark Matter still feasible?}
281: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
282: There is mounting, if not overwhelming, evidence that CDM provides the
283: most accurate description of our Universe. Observations point towards
284: a \LCDM\ Universe comprised of 28\% dark matter, 68\% dark energy, and
285: luminous baryonic matter (i.e. galaxies, stars, gas, and dust) at a
286: mere 4\% (cf. Spergel~\ea 2003). This so-called ``concordance model''
287: induces hierarchical structure formation whereby small objects form
288: first and subsequently merge to form progressively larger objects
289: (e.g. White \& Rees 1978; Davis \ea 1985). Hence, galaxies and galaxy
290: clusters are constantly fed by an accretion stream of smaller entities
291: starting to orbit within the encompassing dark matter potential of the
292: host. While generally successful, the \LCDM\ model does face several
293: problems, one such problem actually being the prediction that
294: one-to-two orders of magnitude more satellite galaxies should be
295: orbiting within galactic halos than are actually observed (Klypin~\ea
296: 1999; Moore~\ea 1999).
297:
298: However, there are also indications that the CDM model is in fact
299: correct and does \textit{not} have a problem with an overabundant
300: population of satellite galaxies. For instance, Benson~\ea (2002)
301: carried out a semi-analytical study of satellites in the Local Group
302: and found that an earlier epoch of reionisation was sufficient to
303: suppress star formation in many of the subhalos and thus produce a
304: significant population of ``dark galaxies''.
305:
306: Therefore, if the CDM model is in fact correct and the (overabundant)
307: population of (dark) satellites predicted by it really does exist, it
308: is imperative to understand the discrepancy by investigating the
309: orbital evolution of these objects and their deviation from the
310: background dark matter distribution.
311:
312: \subsubsection*{The story, so far}
313: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
314: To date, typical satellite properties such as orbital parameters and
315: mass loss under the influence of the host halo have primarily been
316: investigated using \textit{static} potentials for the dark matter host
317: halo (Johnston \ea 1996; Hayashi \ea 2003). We stress that each of
318: these studies have provided invaluable insights into the physical
319: processes involved in satellite disruption; our goal is to augment
320: those studies by relaxing the assumption of a static host potential
321: as, in practice, realistic dark matter halos are neither static nor
322: spherically symmetric.
323:
324: \subsubsection*{The story continues}
325: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
326: The work presented here is based upon a set of numerical simulations
327: of structure formation within said concordance model, analysing in
328: detail the temporal and spatial properties of satellite galaxies
329: residing within host dark matter halos that formed fully
330: self-consistently within a cosmological framework. We focus on
331: interactions between satellite galaxies orbiting within a larger dark
332: matter halo and especially if there is a relation between mutual
333: interplay and distance to the host. The outline of the paper is as
334: follows. In \Sec{Identify} we present our new halo finding algorithms
335: based upon the \nbody\ code \mlapm. We then apply it to our set of
336: eight cosmological dark matter halos in
337: \Sec{Application} with a summary of ours results given in
338: \Sec{Summary}.
339:
340:
341:
342:
343: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
344: \section{Identifying Satellite Galaxies}\label{Identify}
345: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
346: \subsubsection*{Cosmological Simulations}
347: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
348: Over the last decades great advancements have been made in the
349: development of \nbody\ codes. We have seen the rise of tree based
350: gravity solvers (Barnes and Hut 1986), mesh based techniques
351: (Klypin~\& Shandarin 1983), and combinations of direction summation
352: techniques and grid based Poisson solvers (Efstathiou \ea 1985).
353: However, simulating the Universe in a computer and producing the data
354: is only the first step in a long journey; the purpose of these codes
355: is their predictive power, thus the ensembles of millions of dark
356: matter particles used with such (dissipationless) \nbody\ codes need
357: to be interpreted and then compared to the observable Universe. This
358: task requires analysis tools to map the phase-space, which is being
359: sampled by the particles, back to "real" objects in the Universe, the
360: traditional way has been through the use of "halo finders".
361:
362: \subsubsection*{Identifying Dark Matter Halos}
363: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
364: Halo finders mine the \nbody\ data to find locally overdense
365: gravitationally bound systems. Under the assumption that all galaxies
366: and galaxies clusters are centered about local over-density peaks in
367: the dark matter density field they are usually found just using
368: spatial information of the particle distribution. To identify objects
369: in this fashion, the halo finder is required in some way to reproduce
370: the work of the \nbody\ solver in the calculation of the density field
371: or the location of its peaks. The major limitation, however, will
372: always be the appropriate reconstruction of the density
373: field. Normally this task is performed \textit{after} the simulation
374: has finished using an independent method to derive a) the density
375: field and b) to smooth it on a certain scale. With that in mind, we
376: are using a new method for identifying gravitationally bound objects
377: that utilizes the adaptive meshes of the open source \nbody\ code
378: \mlapm\footnote{\mlapm\ can be downloaded from the webpage
379: \texttt{http://astronomy.swin.edu.au/MLAPM}}(Knebe, Green~\& Binney
380: 2001). It is called \mhf\ (\mlapm's Halo Finder) and naturally works
381: on-the-fly, but has also been adapted to deal with single outputs of
382: any \nbody\ code. However, in order to understand the functionality of
383: \mhf\ it is important to gain insight into the mode of operation of
384: \mlapm\ first.
385:
386:
387: \begin{figure}
388: \begin{center}
389: \psfig{file=eps/MLAPMref.eps,width=\hsize}
390: \caption{\mlapm\ at work. The upper panels show a sample cosmological
391: \LCDM\ simulation with the lower panels zooming into the
392: marked region. In the left panels the particle positions are
393: plotted whereas the right panels are indicating the
394: (adaptive) grid points used to solve the governing equations
395: of motion. The circle in the lower right panel highlights
396: substructure being picked up by the finest refinement grid.}
397: \label{MLAPMref} % for cross-references
398: \end{center}
399: \end{figure}
400:
401: \begin{figure}
402: \begin{center}
403: \psfig{file=eps/Orbits.eps,width=\hsize}
404: \caption{Some sample orbits of satellite galaxies within our set
405: of dark matter host halos. We can clearly see how well
406: we trace the orbits and follow the tidal disruption
407: of the satellites, respectively.}
408: \label{Orbits} % for cross-references
409: \end{center}
410: \end{figure}
411:
412:
413: \subsubsection*{\mlapm's Mode of Operation}
414: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
415: \mlapm\ reaches high force resolution by refining high-density regions
416: with an automated refinement algorithm. These adaptive meshes are
417: recursive: refined regions can also be refined, each subsequent
418: refinement having cells that are half the size of the cells in the
419: previous level. This creates a hierarchy of refinement meshes of
420: different resolutions covering regions of interest. The refinement is
421: done cell by cell (individual cells can be refined or de-refined) and
422: meshes are not restricted to have a particular symmetry. The criterion
423: for (de-)refining a cell is simply the number of particles within that
424: cell and a detailed study of the appropriate choice for this number
425: can be found elsewhere (Knebe~\ea 2001). \mlapm's adaptive refinement
426: meshes therefore follow the density distribution by
427: construction. Thus, the grid structure naturally surrounds the
428: (satellite) galaxies as they manifest themselves as over-densities in
429: the underlying background field, an example of which can be viewed in
430: \Fig{MLAPMref} where we show a slice through a sample \LCDM\
431: simulation. In the left panels the actual particle distribution is
432: presented whereas the right panels indicate the adaptive meshes
433: invoked by \mlapm\ to solve Poisson's equation and integrate the
434: equations of motion, respectively. In the lower right panel the white
435: circle highlights the ability of \mlapm'grid to locate substructure:
436: only on the finest refinement level it becomes apparent that the
437: massive galaxy cluster in fact has two centres which is a mere
438: reflection of the fact it recently underwent a major merger with the
439: two progenitors still not fully coalesced yet. The advantage of
440: reconstructing and using these adaptive grids to identify prospective
441: halo centres is that they naturally follow the density field with the
442: \textit{exact} accuracy of the \nbody\ code.
443:
444: \subsubsection*{\mhf\ (\mlapm's-Halo-Finder)}
445: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
446: In \Fig{MLAPMref} we have seen the capability of \mlapm\ to localise
447: local overdensity peaks in cosmological simulations of structure
448: formation. But this is just the first step to identifying
449: gravitationally bound objects. To actually locate dark matter halos
450: within the simulation data we build a register of positions of the
451: peaks in the density field from the full adaptive grid structure
452: invoked by \mlapm\ using the same refinement criterion as for the
453: original runs; we build a list of "potential centers". To do this we
454: restructure the hierarchy of nested isolated \mlapm\ grids into a
455: "grid tree" storing the centre of the densest grid in the end of each
456: branch. For each of these potential centres we step out in radial bins
457: until the overdensity (measured in terms of the cosmological
458: background density) drops below the virial value set by the background
459: cosmological model, i.e. $\Delta_{\rm vir}=340$ for \LCDM\ at redshift
460: $z=0$. This defines the virial radius \Rvir\ and provides us with a
461: list of particles associated with that dark matter halo.
462:
463: We then need to prune that list and remove (in an iterative procedure)
464: all gravitationally unbound particles, respectively. Starting with
465: the potential centre again, we calculate the kinetic and potential
466: energy for each individual particle in the respective reference frame
467: and all particles faster than two times the escape velocity are
468: removed from the halo. We then recalculate the centre (as well as the
469: virial radius) and proceed through the process again. This iteration
470: stops once no further particles are removed or if there are fewer than
471: eight particles left in which case the potential centre will be
472: removed from the halo list completely.
473:
474: In the end we are left with not only a list of appropriate halo
475: positions but we also derived canonical properties for all credible
476: objects, e.g. virial radius, virial mass, velocity dispersion, density
477: profile, etc. A more elaborate description of our technique can be
478: found elsewhere though (Gill, Knebe~\& Gibson 2004a).
479:
480:
481: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
482: \section{Quantifying Interactions in simulated Galaxy Clusters} \label{Application}
483: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
484:
485: \subsubsection*{The Dark Matter Host Halos}
486: %-------------------------------------------
487: We created a set of eight high-resolution galaxy clusters each
488: consisting of order more than a million dark matter particles. These
489: clusters formed in dissipationless \nbody\ simulations of the
490: so-called "concordance" (\LCDM) cosmology ($\Omega_0 =
491: 0.3,\Omega_\lambda = 0.7, \Omega_b h^2 = 0.022, h = 0.7, \sigma_8 =
492: 0.9$). The runs have a mass resolution of $m_p = 1.6 \times
493: 10^{8}$\hMsun\ and achieved a force resolution of $\approx$2\hkpc\
494: allowing us to resolve the host halos down to about the central 0.25\%
495: of their virial radii \Rvir.
496:
497: The halos were specifically selected to investigate the evolution of
498: satellite galaxies and its debris in an unbiased sample of host halos
499: thus analysing the influence of environment in the evolution of such
500: systems. To achieve this goal high temporal information was required
501: to track the development of the satellites. We therefore stored 17
502: outputs from $z=2.5$ to $z=0.5$ equally spaced with $\Delta t \approx
503: 0.35$~Gyrs. From $z=0.5$ to $z=0$ we have 30 outputs spaced $\Delta t
504: \approx 0.17$~Gyrs. A summary of the eight host halos is presented in
505: Table~\ref{HaloDetails}.
506:
507: The quality of our halo finder and our data, respectively, can be
508: viewed in \Fig{Orbits}. There we show the orbits of four sample
509: satellite galaxies orbiting within their respective host halo. This
510: Figure nicely demonstrates how we are very accurately tracking the
511: orbits of the satellites within the area of trade of the host
512: halos. In a companion paper (Gill~\ea 2004) we are presenting a
513: thorough analysis of the dynamics of these satellite galaxies. There
514: we also present the number distribution of orbits of the substructure
515: population which peaks at about 1--2 orbits with a tail extending to
516: as many as 5 orbits in the older systems. However, in this study we
517: like to focus on one particular aspect, namely satellite-satellite
518: encounters.
519:
520: \subsubsection*{Quantifying Encounters}
521: %-------------------------------------------
522: As a first order approximation for quantifying encounters between
523: substructure galaxies we calculated the tidal radius of a given
524: satellite \textit{induced by one of the other satellites}. This means
525: that the tidal radius is defined to be the radius where the
526: gravitational effects of the companion satellite are greater than its
527: self-gravity. When approximating both satellites as point masses and
528: maintaining that the mean density within the satellite has to be three
529: times the mean density of the "perturber" at distance $D$ (Jacobi
530: limit) the definition for tidal radius reads as follows
531:
532: \begin{equation}\label{TidalRad}
533: r_{\rm tidal} = \left( \frac{m}{3M} \right)^{\frac{1}{3}} D \ ,
534: \end{equation}
535:
536: \noindent
537: where $m$ is the mass of the actual satellite and $M$ is the mass of
538: the perturbing satellite at distance $D$.
539:
540:
541: Whenever the tidal radius becomes smaller than the virial
542: radius\footnote{We are tracking each satellite galaxy individually
543: from the formation time of the host halo using its initial particle
544: content and hence we are in the unique position to accurately
545: calculate its virial radius as the radius where the mean averaged
546: density (measured in terms of the cosmological background density
547: $\rho_b$) drops below $\Delta_{\rm vir}(z)$.} of the satellite we
548: increased a counter for that particular satellite. This counter now
549: keeps track of the number of (perturbing) interactions with companion
550: satellite galaxies. As some of the satellites may have had more
551: interactions simply because they spent more time orbiting the host we
552: are normalising the number of encounters by the number of orbits for
553: each individual satellite. The distribution of this (normalised)
554: counter is presented in \Fig{Nenc}. The well pronounced peak at zero
555: encounters shows that in most cases the interactions between
556: satellites is negligible. However, we also observe that (in our
557: simplistic treatment for satellite-satellite interactions) we do find
558: as many as 3-4 encounters per orbit for individual satellites. This,
559: in fact, indicates that with sufficient (spatial) resolution (as it is
560: the case with our data) one is able to decipher the influence of the
561: dominant host halo from the (more minor) interactions with the
562: companion satellite galaxies. We, however, leave a detailed analysis
563: of this phenomenon to a companion paper (Gill, Knebe~\& Gibson 2004b),
564: where we individually select satellite galaxies and resimulate them in
565: static and evolving analytic host potentials as opposed to their
566: evolution in the live potential used for this study.
567:
568: We complement \Fig{Nenc} with \Table{Encounters} where we give the
569: percentage of satellites that had one or more encounters per orbit.
570: The average percentage amounts to 30\% of the whole substructure
571: population. We also observe a clear trend for the interactions to
572: become more prominent in younger systems. This is basically a reflection
573: of the fact that the younger systems are still in the process of
574: digesting their last major merger and have not reached an equilibrium
575: state yet, respectively.
576:
577:
578: \subsubsection*{Relations to Observations}
579: %--------------------------------------------
580: If we now assume that such interactions might be held responsible for
581: star formation bursts, i.e. if encounters trigger star formation, it
582: raises the question whether we can explain the observed correlation
583: between star formation activity in the Local Group Dwarfs and
584: distance to the centre of the Milky Way. Van den Bergh (1994), for
585: instance, reported that Dwarf spheroidals located close to the Galaxy
586: only experienced star formation early in their lifetimes. Dwarf
587: spheroidals at intermediate distances underwent significant star
588: formation more recently whereas the most distant ones do show ongoing
589: star formation at the present time. Do encounters with other
590: satellites trigger star formation bursts? To this extent we present
591: the relation between the number of encounters (per orbit) as a
592: function of distance to the centre of the host at redshift $z=0$. The
593: result can be viewed in \Fig{EncDist}. Unfortunately we do not
594: observe a clear trend for all our halos, even though most of them
595: actually show the reverse correlation, namely the closer a satellite
596: to the host galaxy the more encounters with other substructure. This
597: relation is even more prominent when not normalising by the number of
598: orbits. Only halo~\#7 does show a trend that agrees with the
599: observational finding for star formation activity and distance to the
600: centre, even though we show in Gill~\ea (2004) that halo~\#7 does
601: otherwise have no outstanding differences to the other halos. Anyway,
602: as we see in Gill, Knebe~\& Gibson (2004a) the radial satellite
603: density distribution roughly declines like $\rho_{\rm sat} \propto
604: r^{-2}$ and hence the mild (anti-)correlation between number of
605: encounters and distance can be interpreted as a ``volume effect'':
606: closer to the centre of the host lives approximately the same number
607: of satellites in a spherical shell as farther out, but as the volume
608: of that shell is smaller it is more likely for the satellites to
609: interact.
610:
611: \begin{figure}
612: \begin{center}
613: \psfig{file=eps/Nenc.eps,width=\hsize}
614: \caption{Distribution of number of encounters for all satellite galaxies
615: more massive than 10$^{10}$\hMsun\ at redshift $z=0$.}
616: \label{Nenc} % for cross-references
617: \end{center}
618: \end{figure}
619:
620: \begin{figure}
621: \begin{center}
622: \psfig{file=eps/EncDist.eps,width=\hsize}
623: \caption{Encounters per orbit as a function of distance to
624: the host halo's centre for redshift $z=0$.}
625: \label{EncDist} % for cross-references
626: \end{center}
627: \end{figure}
628:
629:
630:
631: \begin{table}
632: \begin{center}
633: \caption{Properties of the eight dark matter host halos. Distances are
634: measured in \hMpc, velocities in \kms, masses in
635: 10$^{14}$\hMsun, and the age in Gyrs. We applied a mass-cut
636: of $M>10^{10}$\hMsun\ (100 particles) which explains the rather
637: 'low' number for $N_{\rm sat}(<\!R_{\rm vir})$.}
638: \label{HaloDetails}
639: \begin{tabular}{cccccc}\hline
640: Halo & \Rvir & \Mvir & \zform & age &
641: $N_{\rm sat}(<\!R_{\rm vir})$ \\
642:
643: \hline \hline
644: \# 1 & 1.34 & 2.87 & 1.16 & 8.30 & 158 \\
645: \# 2 & 1.06 & 1.42 & 0.96 & 7.55 & 63 \\
646: \# 3 & 1.08 & 1.48 & 0.87 & 7.16 & 87 \\
647: \# 4 & 0.98 & 1.10 & 0.85 & 7.07 & 57 \\
648: \# 5 & 1.35 & 2.91 & 0.65 & 6.01 & 175 \\
649: \# 6 & 1.05 & 1.37 & 0.65 & 6.01 & 85 \\
650: \# 7 & 1.01 & 1.21 & 0.43 & 4.52 & 59 \\
651: \# 8 & 1.38 & 3.08 & 0.30 & 3.42 & 251 \\
652: \end{tabular}
653: \end{center}
654: \end{table}
655:
656:
657: \begin{table}
658: \begin{center}
659: \caption{Percentage of satellites that had one or more
660: encounters per orbit.}
661: \label{Encounters}
662: \begin{tabular}{cc}\hline
663: Halo & percentage \\
664:
665: \hline \hline
666: \# 1 & 14 \\
667: \# 2 & 18 \\
668: \# 3 & 12 \\
669: \# 4 & 31 \\
670: \# 5 & 27 \\
671: \# 6 & 22 \\
672: \# 7 & 58 \\
673: \# 8 & 58 \\
674: \end{tabular}
675: \end{center}
676: \end{table}
677: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
678: \section{Summary} \label{Summary}
679: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
680: We used a set of eight high-resolution cosmological simulations to
681: investigate and quantify interactions between satellite galaxies
682: orbiting within a common dark matter halo. Using our definition for
683: encounter, which is based upon the mutually induced tidal radius, we
684: showed that on average 30\% of the substructure population had had
685: more than one encounter per orbit with another satellite galaxy
686: orbiting within the same host halo. There is, however, a clear trend
687: for interactions to be more common in young galaxy clusters. We
688: furthermore showed that satellite galaxies closer to the centre of the
689: host halo had had more interactions with companion satellites, not
690: because they simply orbited for longer in the underlying host
691: potential but most likely because of the universal radial distribution
692: of satellite galaxies found in cosmological dark matter halos
693: (Gill~\ea 2004). Even though satellite-satellite interactions are
694: unimportant for the majority of satellite galaxies, there exists a
695: sub-population for which this needs to be investigated in more detail
696: and more carefully, respectively.
697:
698: We also noted that there is a degeneracy between the influence of the
699: host halo and the interactions with the companion satellites which can
700: only be disentangled with an appropriate resolution for both the
701: actual \nbody-simulation and the halo finding technique. We therefore
702: applied a new method for identifying gravitationally bound objects
703: in cosmological \nbody\ simulations. This new technique is based upon
704: the adaptive grid structures of the open source adaptive mesh
705: refinement code \mlapm\ (Knebe, Green~\& Binney 2001). The halo finder
706: is called \mhf\ and acts on the same accuracy level as the actual
707: simulation. A more thorough study of the functionality of \mhf\ is
708: presented in Gill, Knebe~\& Gibson (2004a). A detailed analysis
709: of the degeneracy between influence of the host halo and interactions
710: with companion satellites can be found in a companion paper, too
711: (Gill, Knebe~\& Gibson 2004b).
712:
713: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
714: \section*{Acknowledgments}
715: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
716: The simulations presented in this paper were carried out on the
717: Beowulf cluster at the Centre for Astrophysics \& Supercomputing,
718: Swinburne University.
719:
720: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
721: \section*{References}
722: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
723:
724: % PASA uses the same conventions as ApJ for journal abbreviations. Sample
725: % references are as follows.
726: % Please follow the same format for your references.
727:
728: \reference Barnes J. E., Hut P., 1986, Nature 324, 446
729: \reference Benson A.J., Frenk C.S., Lacey C.G., Baugh C.M., Cole S., 2002, MNRAS 333, 177
730: \reference Bertschinger E., 1998, ARA\&A 36, 599
731: \reference Davis M., Efstathiou G., Frenk C. S., \& White S. D. M., 1985, ApJ 292, 371
732: \reference Efstathiou G., Davis M., White S.D.M., Frenk C.S., 1985, ApJS 57, 241
733: \reference Gill S.P.D., Knebe A., Gibson B.K., 2004, MNRAS submitted
734: \reference Gill S.P.D., Knebe A., Gibson B.K., 2004, in preparation
735: \reference Gill S.P.D., Knebe A., Gibson B.K., Dopita M.A., 2004, MNRAS submitted
736: \reference Hayashi E., Navarro J., Taylor J., Stadel J., Quinn T., 2003, ApJ 584, 541
737: \reference Ibata R.A., Lewis G.F., 1998, ApJ 500, 575
738: \reference Johnston K., Hernquist L., Bolte M., 1996, ApJ 465, 278
739: \reference Klypin A.A., Shandarin, S. F., 1983, MNRAS 204, 891
740: \reference Klypin A., Kravtsov A., Valenzuela O., Prada F., 1999, ApJ 522, 82
741: \reference Knebe A., Green A., Binney J., 2001, MNRAS 325, 845
742: \reference Moore B., Ghigna S., Governato F., Lake G., Quinn T., Stadel J.,
743: Tozzi P., 1999, ApJL 524, 19
744: \reference Spergel D. N., Verde L., Peiris H. V., Komatsu E., Nolta M. R.,
745: Bennett C. L., Halpern M., Hinshaw G., Jarosik N., Kogut A.,
746: Limon M., Meyer S. S., Page L., Tucker G. S., Weiland J. L.,
747: Wollack E., Wright E. L., 2003, ApJS 148, 175
748: \reference van den Bergh S., 1994, ApJ 428, 617
749: \reference White S. D. M., \& Rees M., 1978, MNRAS 183, 341
750: \reference Zhao H., 1998, ApJL 500, 149
751:
752:
753: \end{document}
754: